How do you convert the x y of a tab file into lat long then expose the lat long as attributes into a excel table? (I have point and area features)

david_r wrote:

Lots of ways, but here's a suggestion

1. CsmapReprojector to go from X/Y to Lat/Long
2. CoordinateExtrator to extract the coordinates as attributes
3. Excel writer

 

First off thanks for your quick reply, I now feel I'm almost there ... but not quite. My original data coordinate system is BritishNationalGrid. When I applie either LL84 or LL_WGS84 two things happen in the repojector.

 

1. The lat longs seem to be getting written out the wrong way round e.g. -1.5747253716770673,53.681833979608278 rather than 53.681833979608278, -1.5747253716770673

 

2. The location is consistency around 1.4m offset in a north west direction. E.g. true location using X Y is 428188.06, 420679.96, when processed through the workbench the lat long is 53.681833979608278, -1.5747253716770673 the X Y of this is 428186.96 420680.97

 

 

Can you suggest how to correct the offset which my workbench is creating and is there a way to switch the lat longs round?

 

 

I'm new to this type of functionality within FME so please bare with me. Thanks

 

 

jez wrote:

 

First off thanks for your quick reply, I now feel I'm almost there ... but not quite. My original data coordinate system is BritishNationalGrid. When I applie either LL84 or LL_WGS84 two things happen in the repojector.

 

1. The lat longs seem to be getting written out the wrong way round e.g. -1.5747253716770673,53.681833979608278 rather than 53.681833979608278, -1.5747253716770673

 

2. The location is consistency around 1.4m offset in a north west direction. E.g. true location using X Y is 428188.06, 420679.96, when processed through the workbench the lat long is 53.681833979608278, -1.5747253716770673 the X Y of this is 428186.96 420680.97

 

 

Can you suggest how to correct the offset which my workbench is creating and is there a way to switch the lat longs round?

 

 

I'm new to this type of functionality within FME so please bare with me. Thanks

 

 

Unfortunately I know nothing about the British National Grid, hopefully someone else (@1spatialdave?) with some local knowledge can chime in.

 

As for the switched lat/long values, how are you naming them in the CoordinateExtractor and how are you writing them out? Is it possible that you've switched them around and it's just a question of ordering them properly in the Writer?

david_r wrote:

Unfortunately I know nothing about the British National Grid, hopefully someone else (@1spatialdave?) with some local knowledge can chime in.

 

As for the switched lat/long values, how are you naming them in the CoordinateExtractor and how are you writing them out? Is it possible that you've switched them around and it's just a question of ordering them properly in the Writer?

not sure why the coordinates appear to be swapping around, I haven't noticed that before, other than getting it confused myself!

 

 

the offset is a common issue though and is (often) due to the default projection transformation used to switch between the OSGB36 (BNG) datum and the WGS84 (LL) datum.

 

The default is nearly always pretty rough.

 

the best, most accurate, transformer to use is the GridInquest transformer that uses the OSTN02 grid transformation. If you're not down to the cm level though, you could use the CSMAPreprojector, picking the 7 param/petroleum transformation, or the ESRI reprojector, if you have ArcGIS desktop installed - again you need to pick the 7param/petroleum transformation.

 

 

the gridInquest Transformer outputs the coordinates in ETRS89, not WGS84 however, but the 2 are often seen as synonymous.

 

 

another gotcha with the gridinquest transformer is that the grid, containing the data that is used in the transformation, only extends 10Km offshore, so if you have offshore data use the other tools. I've created a wrapper for this in the fme store, which will direct you to the right transformer.

 

 

ESRI have published some good blogs on this subject.

 

 

nrich wrote:

not sure why the coordinates appear to be swapping around, I haven't noticed that before, other than getting it confused myself!

 

 

the offset is a common issue though and is (often) due to the default projection transformation used to switch between the OSGB36 (BNG) datum and the WGS84 (LL) datum.

 

The default is nearly always pretty rough.

 

the best, most accurate, transformer to use is the GridInquest transformer that uses the OSTN02 grid transformation. If you're not down to the cm level though, you could use the CSMAPreprojector, picking the 7 param/petroleum transformation, or the ESRI reprojector, if you have ArcGIS desktop installed - again you need to pick the 7param/petroleum transformation.

 

 

the gridInquest Transformer outputs the coordinates in ETRS89, not WGS84 however, but the 2 are often seen as synonymous.

 

 

another gotcha with the gridinquest transformer is that the grid, containing the data that is used in the transformation, only extends 10Km offshore, so if you have offshore data use the other tools. I've created a wrapper for this in the fme store, which will direct you to the right transformer.

 

 

ESRI have published some good blogs on this subject.

 

 

Thanks chiming in about the British national grid, great info!

jez wrote:

 

First off thanks for your quick reply, I now feel I'm almost there ... but not quite. My original data coordinate system is BritishNationalGrid. When I applie either LL84 or LL_WGS84 two things happen in the repojector.

 

1. The lat longs seem to be getting written out the wrong way round e.g. -1.5747253716770673,53.681833979608278 rather than 53.681833979608278, -1.5747253716770673

 

2. The location is consistency around 1.4m offset in a north west direction. E.g. true location using X Y is 428188.06, 420679.96, when processed through the workbench the lat long is 53.681833979608278, -1.5747253716770673 the X Y of this is 428186.96 420680.97

 

 

Can you suggest how to correct the offset which my workbench is creating and is there a way to switch the lat longs round?

 

 

I'm new to this type of functionality within FME so please bare with me. Thanks

 

 

 

Are you sure you've calculated that offset right, the offset I see when reprojecting back to BNG is 1.4558657187077331e-006